This invention relates to the art of brazing and more particularly to a method of brazing metal members together by the use of a precision insert.
The prior art in brazing, particularly in tube to tube brazing, suffers from the specific problems of low and variable contact area along the length of abutting brazed tube joints, lack of inspectability of the brazed joints, and the presence of large braze fillets at the bottom end of the joints. The variable contact area problem arose from the practice of not having the inserts precisely conforming to the curvature of the members to be joined which resulted in a weak bond. The non-inspectability of the joints was due to the fact that once the braze inserts were in place it became extremely difficult to ascertain the closeness of the fit which also resulted in a variable contact area problem. In addition, the difficulty in determining the proper amount of braze metal to be used usually resulted in the application of an excess amount as a precautionary measure. This excess amount tended to run off the members to which it was applied thus eroding the members to be joined and creating braze fillets in inappropriate locations resulting in high stress concentrations thereby reducing the reliability of the joint.
An example of the prior art, is a tube to tube brazing process wherein at least two tubes are placed along side each other and wire inserts are tightly placed in the V-like gap along the length of the abutting tubes. These wires may be welded in place or held in position by a number of appropriate means well known in the art. A braze material is then applied to the wires and the entire assembly is heated causing the braze material to flow between the wires and the tubes. Cooling of the assembly results in the formation of a bond among the wires and tubes. In some applications due to the large size of the assemblies, the assemblies must be heated in a vertical position because only vertical furnaces are large enough to accommodate these large assemblies. In this process, experience has shown that the space between the wire inserts is too large to hold the braze metal against the force of gravity. This causes gaps of various sizes along the tubes where the braze metal flows out. Because some of these gaps are larger than others, the braze material does not uniformly fill these gaps which results in low and variable contact areas which in turn results in a poor bond. In an attempt to compensate for this deficiency, an excess amount of braze material is used in the hope of filling the gaps. However, even the use of an excess amount of braze material will not fill these gaps. Further, when a vertical furnace is used, this excess braze metal tends to run down the assembly eroding the wires and tubes and forming large braze fillets at inappropriate locations which result in high stress concentrations. In addition, the wire inserts prevent inspection of the joint.
In U.S. Pat. No. 2,182,796, to Deming, issued Dec. 12, 1939, there is described a welding process wherein the edges of the members to be joined are heated along with the heating of a filler metal so that when pressed together surface fusion produces a welded seam joining the members. The patent also describes the use of a brazing metal in this process. This patented method teaches that the process is most successful when the lateral surfaces of the filler rod are heated to such an extent as to supply an excess of molten metal which flows into and fills vacant areas between the filler rod and the members to be joined. This process exemplifies the prior art's deficiencies in producing hot flowing metal which erodes surfaces thus producing defects in the metals to be joined. This situation is unacceptable where high quality assurance levels of the metal surfaces are required. The patent to Deming also teaches that although conformity of the insert to the surfaces of the metals is desirable, it is not necessary because under enough pressure the insert when heated can be made to conform to the surfaces of the members to be joined. However, in practice the reliability of this concept, especially on vertically oriented welding processes, has proven to be suspect. This is particularly the case where the surfaces of the members to be joined are not susceptible to being beveled as the Deming patent suggests. Although the recommendation of the Deming patent that the pieces to be joined be placed a short distance apart may be advisable in some circumstances, this recommendation may not be advisable in situations where the pieces to be joined must be in contact without the aid of the joining medium.
In U.S. Pat. No. 2,719,210, to Chapman, issued Sept. 27, 1955, there is described a method of welding together thin walled tubes from a single side. This method utilizes a backing strip in conjunction with a filler rod such that when the filler rod is heated and pressure is applied to the filler rod, the filler rod is pressed against the backing strip, forming a fin which fuses with each tube. Although the patent of Chapman describes a method that may be used where there is access to both sides of the tubes, it does not solve the problem of joining such tubes where access, being limited, prohibits the use of such a backing strip. In addition, the patent suggests that if conditions are not precise, the backing strip may become bonded to the joint.
The patent to Chapman teaches that the fin thickness should be minimized which produces deficiencies in application wherein it is important to provide a joint substantially coextensive with the diameter of the tubes such as in current carrying tubes of electromagnetic pumps. In those types of applications it is important to provide a joint substantially coextensive with the tube diameter to enable a current, 4 uniform across the tube diameter, to be carried across a series of tubes. Also, the thin fin attachment does not provide a large cross-sectional attachment area which in turn results in higher stress loads on those attachments.
The Chapman patent also acknowledges that excessive heating of the tubes may result in the burning of holes in the tube wall. To circumvent this problem Chapman applies heat only to the filler rod that produces the fin attachment. However, the Chapman patent does not solve the problem of preventing defects in the tube wall where it is advantageous to heat the entire assembly.
Methods of joining tubes by the use of stamped sheet inserts are known in the prior art but these methods use direct welding of the tubes to the inserts which may produce defects in the tubes. Also, various other methods utilizing inserts of braze metal are known, but these methods do not utilize precision inserts.